October 16, 2019

2499 words 12 mins read

Paper Group NAWR 4

Paper Group NAWR 4

Chengyu Cloze Test. Beyond Generic Summarization: A Multi-faceted Hierarchical Summarization Corpus of Large Heterogeneous Data. Binarized Attributed Network Embedding. MGAD: Multilingual Generation of Analogy Datasets. Sudachi: a Japanese Tokenizer for Business. Parallel Feature Pyramid Network for Object Detection. Crowd Counting via Adversarial …

Chengyu Cloze Test

Title Chengyu Cloze Test
Authors Zhiying Jiang, Boliang Zhang, Lifu Huang, Heng Ji
Abstract We present a neural recommendation model for Chengyu, which is a special type of Chinese idiom. Given a query, which is a sentence with an empty slot where the Chengyu is taken out, our model will recommend the best Chengyu candidate that best fits the slot context. The main challenge lies in that the literal meaning of a Chengyu is usually very different from it{'}s figurative meaning. We propose a new neural approach to leverage the definition of each Chengyu and incorporate it as background knowledge. Experiments on both Chengyu cloze test and coherence checking in college entrance exams show that our system achieves 89.5{%} accuracy on cloze test and outperforms human subjects who attended competitive universities in China. We will make all of our data sets and resources publicly available as a new benchmark for research purposes.
Tasks
Published 2018-06-01
URL https://www.aclweb.org/anthology/W18-0516/
PDF https://www.aclweb.org/anthology/W18-0516
PWC https://paperswithcode.com/paper/chengyu-cloze-test
Repo https://github.com/bazingagin/chengyu_data
Framework none

Beyond Generic Summarization: A Multi-faceted Hierarchical Summarization Corpus of Large Heterogeneous Data

Title Beyond Generic Summarization: A Multi-faceted Hierarchical Summarization Corpus of Large Heterogeneous Data
Authors Christopher Tauchmann, Thomas Arnold, Andreas Hanselowski, Christian M. Meyer, Margot Mieskes
Abstract
Tasks Document Summarization, Multi-Document Summarization
Published 2018-05-01
URL https://www.aclweb.org/anthology/L18-1503/
PDF https://www.aclweb.org/anthology/L18-1503
PWC https://paperswithcode.com/paper/beyond-generic-summarization-a-multi-faceted
Repo https://github.com/AIPHES/HierarchicalSummarization
Framework none

Binarized Attributed Network Embedding

Title Binarized Attributed Network Embedding
Authors Hong Yang, Shirui Pan, Peng Zhang, Ling Chen, Defu Lian, Chengqi Zhang
Abstract An implementation of “Binarized Attributed Network Embedding”. Attributed network embedding enables joint representation learning of node links and attributes. Existing attributed network embedding models are designed in continuous Euclidean spaces which often introduce data redundancy and impose challenges to storage and computation costs. To this end, we present a Binarized Attributed Network Embedding model (BANE for short) to learn binary node representation. Specifically, we define a new Weisfeiler-Lehman proximity matrix to capture data dependence between node links and attributes by aggregating the information of node attributes and links from neighboring nodes to a given target node in a layer-wise manner. Based on the Weisfeiler-Lehman proximity matrix, we formulate a new Weisfiler-Lehman matrix factorization learning function under the binary node representation constraint. The learning problem is a mixed integer optimization and an efficient cyclic coordinate descent (CCD) algorithm is used as the solution. Node classification and link prediction experiments on real-world datasets show that the proposed BANE model outperforms the state-of-the-art network embedding methods.
Tasks Graph Embedding, Link Prediction, Network Embedding, Node Classification, Representation Learning
Published 2018-10-22
URL https://www.researchgate.net/publication/328688614_Binarized_Attributed_Network_Embedding
PDF https://shiruipan.github.io/pdf/ICDM-18-Yang.pdf
PWC https://paperswithcode.com/paper/binarized-attributed-network-embedding
Repo https://github.com/benedekrozemberczki/karateclub
Framework none

MGAD: Multilingual Generation of Analogy Datasets

Title MGAD: Multilingual Generation of Analogy Datasets
Authors Mostafa Abdou, Artur Kulmizev, Vinit Ravishankar
Abstract
Tasks Machine Translation, Part-Of-Speech Tagging, Word Embeddings
Published 2018-05-01
URL https://www.aclweb.org/anthology/L18-1320/
PDF https://www.aclweb.org/anthology/L18-1320
PWC https://paperswithcode.com/paper/mgad-multilingual-generation-of-analogy
Repo https://github.com/rutrastone/MGAD
Framework none

Sudachi: a Japanese Tokenizer for Business

Title Sudachi: a Japanese Tokenizer for Business
Authors Kazuma Takaoka, Sorami Hisamoto, Noriko Kawahara, Miho Sakamoto, Yoshitaka Uchida, Yuji Matsumoto
Abstract
Tasks Chunking, Lemmatization, Morphological Analysis, Part-Of-Speech Tagging, Tokenization
Published 2018-05-01
URL https://www.aclweb.org/anthology/L18-1355/
PDF https://www.aclweb.org/anthology/L18-1355
PWC https://paperswithcode.com/paper/sudachi-a-japanese-tokenizer-for-business
Repo https://github.com/WorksApplications/Sudachi
Framework none

Parallel Feature Pyramid Network for Object Detection

Title Parallel Feature Pyramid Network for Object Detection
Authors Seung-Wook Kim, Hyong-Keun Kook, Jee-Young Sun, Mun-Cheon Kang, Sung-Jea Ko
Abstract Recently developed object detectors employ a convolutional neural network (CNN) by gradually increasing the number of feature layers with a pyramidal shape instead of using a featurized image pyramid. However, the different abstraction levels of the CNN feature layers often limit the detection performance, especially on small objects. To overcome this limitation, we propose a CNN-based object detection architecture, referred to as a parallel feature pyramid (FP) network (PFPNet), where the FP is constructed by widening the network width instead of increasing the network depth. First, we adopt spatial pyramid pooling and some additional feature transformations to generate a pool of feature maps with different sizes. In PFPNet, the additional feature transformation is performed in parallel, which yields the feature maps with similar levels of semantic abstraction across the scales. We then resize the elements of the feature pool to a uniform size and aggregate their contextual information to generate each level of the final FP. The experimental results confirmed that PFPNet increases the performance of the latest version of the single-shot multi-box detector (SSD) by mAP of 6.4% AP and especially, 7.8% AP_small on the MS-COCO dataset.
Tasks Object Detection
Published 2018-09-01
URL http://openaccess.thecvf.com/content_ECCV_2018/html/Seung-Wook_Kim_Parallel_Feature_Pyramid_ECCV_2018_paper.html
PDF http://openaccess.thecvf.com/content_ECCV_2018/papers/Seung-Wook_Kim_Parallel_Feature_Pyramid_ECCV_2018_paper.pdf
PWC https://paperswithcode.com/paper/parallel-feature-pyramid-network-for-object
Repo https://github.com/Stick-To/PFPNet-tensorflow
Framework tf

Crowd Counting via Adversarial Cross-Scale Consistency Pursuit

Title Crowd Counting via Adversarial Cross-Scale Consistency Pursuit
Authors Zan Shen, Yi Xu, Bingbing Ni, Minsi Wang, Jianguo Hu, Xiaokang Yang
Abstract Crowd counting or density estimation is a challenging task in computer vision due to large scale variations, perspective distortions and serious occlusions, etc. Existing methods generally suffers from two issues: 1) the model averaging effects in multi-scale CNNs induced by the widely adopted L2 regression loss; and 2) inconsistent estimation across different scaled inputs. To explicitly address these issues, we propose a novel crowd counting (density estimation) framework called Adversarial Cross-Scale Consistency Pursuit (ACSCP). On one hand, a U-net structural network is designed to generate density map from input patch, and an adversarial loss is employed to shrink the solution onto a realistic subspace, thus attenuating the blurry effects of density map estimation. On the other hand, we design a novel scale-consistency regularizer which enforces that the sum up of the crowd counts from local patches (i.e., small scale) is coherent with the overall count of their region union (i.e., large scale). The above losses are integrated via a joint training scheme, so as to help boost density estimation performance by further exploring the collaboration between both objectives. Extensive experiments on four benchmarks have well demonstrated the effectiveness of the proposed innovations as well as the superior performance over prior art.
Tasks Crowd Counting, Density Estimation
Published 2018-06-01
URL http://openaccess.thecvf.com/content_cvpr_2018/html/Shen_Crowd_Counting_via_CVPR_2018_paper.html
PDF http://openaccess.thecvf.com/content_cvpr_2018/papers/Shen_Crowd_Counting_via_CVPR_2018_paper.pdf
PWC https://paperswithcode.com/paper/crowd-counting-via-adversarial-cross-scale
Repo https://github.com/RQuispeC/pytorch-ACSCP
Framework pytorch

Multi-Domain Neural Machine Translation with Word-Level Domain Context Discrimination

Title Multi-Domain Neural Machine Translation with Word-Level Domain Context Discrimination
Authors Jiali Zeng, Jinsong Su, Huating Wen, Yang Liu, Jun Xie, Yongjing Yin, Jianqiang Zhao
Abstract With great practical value, the study of Multi-domain Neural Machine Translation (NMT) mainly focuses on using mixed-domain parallel sentences to construct a unified model that allows translation to switch between different domains. Intuitively, words in a sentence are related to its domain to varying degrees, so that they will exert disparate impacts on the multi-domain NMT modeling. Based on this intuition, in this paper, we devote to distinguishing and exploiting word-level domain contexts for multi-domain NMT. To this end, we jointly model NMT with monolingual attention-based domain classification tasks and improve NMT as follows: 1) Based on the sentence representations produced by a domain classifier and an adversarial domain classifier, we generate two gating vectors and use them to construct domain-specific and domain-shared annotations, for later translation predictions via different attention models; 2) We utilize the attention weights derived from target-side domain classifier to adjust the weights of target words in the training objective, enabling domain-related words to have greater impacts during model training. Experimental results on Chinese-English and English-French multi-domain translation tasks demonstrate the effectiveness of the proposed model. Source codes of this paper are available on Github \url{https://github.com/DeepLearnXMU/WDCNMT}.
Tasks Machine Translation
Published 2018-10-01
URL https://www.aclweb.org/anthology/D18-1041/
PDF https://www.aclweb.org/anthology/D18-1041
PWC https://paperswithcode.com/paper/multi-domain-neural-machine-translation-with
Repo https://github.com/DeepLearnXMU/WDCNMT
Framework none

SoMeWeTa: A Part-of-Speech Tagger for German Social Media and Web Texts

Title SoMeWeTa: A Part-of-Speech Tagger for German Social Media and Web Texts
Authors Thomas Proisl
Abstract
Tasks Domain Adaptation, Lemmatization, Machine Translation, Named Entity Recognition, Part-Of-Speech Tagging, Tokenization
Published 2018-05-01
URL https://www.aclweb.org/anthology/L18-1106/
PDF https://www.aclweb.org/anthology/L18-1106
PWC https://paperswithcode.com/paper/someweta-a-part-of-speech-tagger-for-german
Repo https://github.com/tsproisl/SoMeWeTa
Framework none

Revisiting neural relation classification in clinical notes with external information

Title Revisiting neural relation classification in clinical notes with external information
Authors Simon {\v{S}}uster, Madhumita Sushil, Walter Daelemans
Abstract Recently, segment convolutional neural networks have been proposed for end-to-end relation extraction in the clinical domain, achieving results comparable to or outperforming the approaches with heavy manual feature engineering. In this paper, we analyze the errors made by the neural classifier based on confusion matrices, and then investigate three simple extensions to overcome its limitations. We find that including ontological association between drugs and problems, and data-induced association between medical concepts does not reliably improve the performance, but that large gains are obtained by the incorporation of semantic classes to capture relation triggers.
Tasks Feature Engineering, Natural Language Inference, Relation Classification, Relation Extraction, Text Categorization
Published 2018-10-01
URL https://www.aclweb.org/anthology/W18-5603/
PDF https://www.aclweb.org/anthology/W18-5603
PWC https://paperswithcode.com/paper/revisiting-neural-relation-classification-in
Repo https://github.com/SimonSuster/seg_cnn
Framework none

Learning to Blend Photos

Title Learning to Blend Photos
Authors Wei-Chih Hung, Jianming Zhang, Xiaohui Shen, Zhe Lin, Joon-Young Lee, Ming-Hsuan Yang
Abstract Photo blending is a common technique to create aesthetically pleasing artworks by combining multiple photos. However, the process of photo blending is usually time-consuming, and care must be taken in the process of blending, filtering, positioning, and masking each of the source photos. To make photo blending accessible to general public, we propose an efficient approach for automatic photo blending via deep learning. Specifically, given a foreground image and a background image, our proposed method automatically generates a set of blending photos with scores that indicate the aesthetics quality with the proposed quality network and policy network. Experimental results show that the proposed approach can effectively generate high quality blending photos with efficiency.
Tasks
Published 2018-09-01
URL http://openaccess.thecvf.com/content_ECCV_2018/html/Wei-Chih_Hung_Learning_to_Blend_ECCV_2018_paper.html
PDF http://openaccess.thecvf.com/content_ECCV_2018/papers/Wei-Chih_Hung_Learning_to_Blend_ECCV_2018_paper.pdf
PWC https://paperswithcode.com/paper/learning-to-blend-photos
Repo https://github.com/hfslyc/LearnToBlend
Framework none

Entity Linking within a Social Media Platform: A Case Study on Yelp

Title Entity Linking within a Social Media Platform: A Case Study on Yelp
Authors Hongliang Dai, Yangqiu Song, Liwei Qiu, Rijia Liu
Abstract In this paper, we study a new entity linking problem where both the entity mentions and the target entities are within a same social media platform. Compared with traditional entity linking problems that link mentions to a knowledge base, this new problem have less information about the target entities. However, if we can successfully link mentions to entities within a social media platform, we can improve a lot of applications such as comparative study in business intelligence and opinion leader finding. To study this problem, we constructed a dataset called Yelp-EL, where the business mentions in Yelp reviews are linked to their corresponding businesses on the platform. We conducted comprehensive experiments and analysis on this dataset with a learning to rank model that takes different types of features as input, as well as a few state-of-the-art entity linking approaches. Our experimental results show that two types of features that are not available in traditional entity linking: social features and location features, can be very helpful for this task.
Tasks Entity Linking, Learning-To-Rank
Published 2018-10-01
URL https://www.aclweb.org/anthology/D18-1227/
PDF https://www.aclweb.org/anthology/D18-1227
PWC https://paperswithcode.com/paper/entity-linking-within-a-social-media-platform
Repo https://github.com/HKUST-KnowComp/ELWSMP
Framework tf

Improving Named Entity Recognition by Jointly Learning to Disambiguate Morphological Tags

Title Improving Named Entity Recognition by Jointly Learning to Disambiguate Morphological Tags
Authors Onur G{"u}ng{"o}r, Suzan Uskudarli, Tunga G{"u}ng{"o}r
Abstract Previous studies have shown that linguistic features of a word such as possession, genitive or other grammatical cases can be employed in word representations of a named entity recognition (NER) tagger to improve the performance for morphologically rich languages. However, these taggers require external morphological disambiguation (MD) tools to function which are hard to obtain or non-existent for many languages. In this work, we propose a model which alleviates the need for such disambiguators by jointly learning NER and MD taggers in languages for which one can provide a list of candidate morphological analyses. We show that this can be done independent of the morphological annotation schemes, which differ among languages. Our experiments employing three different model architectures that join these two tasks show that joint learning improves NER performance. Furthermore, the morphological disambiguator{'}s performance is shown to be competitive.
Tasks Entity Linking, Knowledge Base Population, Named Entity Recognition, Relation Extraction, Word Embeddings
Published 2018-08-01
URL https://www.aclweb.org/anthology/C18-1177/
PDF https://www.aclweb.org/anthology/C18-1177
PWC https://paperswithcode.com/paper/improving-named-entity-recognition-by-jointly-1
Repo https://github.com/onurgu/joint-ner-and-md-tagger
Framework none

On the Development of a Large Scale Corpus for Native Language Identification

Title On the Development of a Large Scale Corpus for Native Language Identification
Authors Thomas Hudson, Sardar Jaf
Abstract Native Language Identification (NLI) is the task of identifying an author’s native language from their writings in a second language. In this paper, we introduce a new corpus (italki), which is larger than the current corpora. It can be used for training machine learning based systems for classifying and identifying the native language of authors of English text. To examine the usefulness of italki, we evaluate it by using it to train and test some of the well performing NLI systems presented in the 2017 NLI shared task. In this paper, we present some aspects of italki. We show the impact of the variation of italki’s training dataset size of some languages on systems performance. From our empirical finding, we highlight the potential of italki as a large scale corpus for training machine learning classifiers for classifying the native language of authors from their written English text. We obtained promising results that show the potential of italki to improve the performance of current NLI systems. More importantly, we found that training the current NLI systems on italki generalize better than training them on the current corpora.
Tasks Language Identification, Native Language Identification
Published 2018-12-10
URL http://www.ep.liu.se/ecp/article.asp?issue=155&article=012&volume=
PDF http://www.ep.liu.se/ecp/155/012/ecp18155012.pdf
PWC https://paperswithcode.com/paper/on-the-development-of-a-large-scale-corpus
Repo https://github.com/ghomasHudson/italkiCorpus
Framework none

Extracting Relationships by Multi-Domain Matching

Title Extracting Relationships by Multi-Domain Matching
Authors Yitong Li, Michael Murias, Geraldine Dawson, David E. Carlson
Abstract In many biological and medical contexts, we construct a large labeled corpus by aggregating many sources to use in target prediction tasks. Unfortunately, many of the sources may be irrelevant to our target task, so ignoring the structure of the dataset is detrimental. This work proposes a novel approach, the Multiple Domain Matching Network (MDMN), to exploit this structure. MDMN embeds all data into a shared feature space while learning which domains share strong statistical relationships. These relationships are often insightful in their own right, and they allow domains to share strength without interference from irrelevant data. This methodology builds on existing distribution-matching approaches by assuming that source domains are varied and outcomes multi-factorial. Therefore, each domain should only match a relevant subset. Theoretical analysis shows that the proposed approach can have a tighter generalization bound than existing multiple-domain adaptation approaches. Empirically, we show that the proposed methodology handles higher numbers of source domains (up to 21 empirically), and provides state-of-the-art performance on image, text, and multi-channel time series classification, including clinically relevant data of a novel treatment of Autism Spectrum Disorder.
Tasks Domain Adaptation, Time Series, Time Series Classification
Published 2018-12-01
URL http://papers.nips.cc/paper/7913-extracting-relationships-by-multi-domain-matching
PDF http://papers.nips.cc/paper/7913-extracting-relationships-by-multi-domain-matching.pdf
PWC https://paperswithcode.com/paper/extracting-relationships-by-multi-domain
Repo https://github.com/yitong91/Multiple-Domain-Matching-Network
Framework tf
comments powered by Disqus